Anomaly in the SFD Galaxy extinction map and FIR emission from SDSS galaxies

Yasushi Suto Department of Physics, The University of Tokyo& Global scholar, Dept. of Astrophysical Sci., Princeton University

October 24, 2011: TodaiForum andDENET-IAP conference on the accelerating universe

SFD Galaxy extinction map

Ar,SFDSDSS DR7 surveyregion

Ar,SFD [mag]S g

al[d

eg-2

]

after extinctioncorrection

SDSS galaxy surfacenumberdensity

before extinctioncorrection

CollaboratorsThe present talk is based on our on-going collaboration with Toshiya Kashiwagi, A.Taruya, I.Kayo & K.YaharaSee also “The effect of FIR emission from SDSS galaxies on the SFD Galactic extinction map”

K.Yahata, A.Yonehara, Y.Suto, E.L.Turner, T.Broadhurst, & D.P. FinkbeinerPubl.Astron.Soc.Japan 59(2007)205astro-ph/0607098

SFD Galactic extinction map and galaxy surveys

The most fundamental dataset for any astronomical observationTrue large-scale structures revealed only after the extinction correctionIts reliability is of vital importance in precision cosmology

Galactic extinction E(B-V) map (Schlegel, Finkbeiner& Davis 1998; SFD)

SFD procedure to construct the Galactic extinction map

COBE 100μm+240μm maps (0.7deg.pixel)Remove zodiacal light and cosmic infrared backgroundDust temperature map ⇒ temperature-dependent emissivity corrected 100μm map

Calibration of higher angular-resolution IRAS 100μm map (5 arcmin. pixel)Assume

E(B-V)=p×(IRAS 100μm flux)T ; T～1at each region and determine p and T from the dataConvert E(B-V) to Aband adopting SED of ellipticalsand RV=AV/E(B-V)=3.1

ASFD map in SDSS DR7 survey region

3.6x106 galaxies (17.5<r<19.4) in 7270 deg2

from SDSS DR7 photometric catalog

SFD Galaxy extinction map

Ar,SFDSDSS DR7 surveyregion

Ar,SFD [mag]

area

(<A

r,SF

D)[

deg2

]

Estimating Galactic extinction from SDSS galaxy surface density

divide the SDSS DR7 survey area into many small regions according to ASFD

combine those un-contiguous regions into 84 bins with ～100 deg2 eachcompare the galaxy number density Sgal for those bins

SDSS DR7 survey area (color coded according to ASFD)

medium extinction

low extinction

high extinction

Anomalous behavior of SDSS galaxy surface density Sgal as a function of ASFDIf ASFD is perfect, we expect that

Before correction: Sgal should monotonically decrease as a function of ASFD

After correction: Sgal should be constant

OK for ASFD >0.1, but quite the opposite for ASFD<0.1

～70% of the SDSS survey area has ASFD <0.1 !

First pointed out by Yahata et al. (2007) for DR4, and confirmed by Kashiwagi (2011) for DR7

Ar,SFD [mag]

S gal

[deg

-2]

before correction

after correction

DR7

Origin of the anomalyASFD is estimated assuming that the reddening is proportional to the FIR emission flux (100μm)

the anomaly indicates the positive correlation between galaxy surface density and the FIR flux at least where the real extinction is small

100μm flux = Galactic dust + galaxiescontamination by the FIR emission from galaxies proposed by Yahata et al. (2007)

Mock simulations to test the hypothesisDistribute random particles over the DR7 region

the same number and the same r-band magnitude distributions as SDSS galaxies

Assign 100μm flux to each particle sampled from the log-normal distribution of L100μm/Lr

Add those 100μm flux to the original SFD map and compute the extinction

at each pixel (5’x5’)Compute particle surface density as a function of at each pixel

mockSFDSFD AAA Δ+⇒

mockSFD AA Δ+

Observed correlation between L100μm and Lr

Distribution function of galaxy luminosities in IRAS 100μm vs. SDSS r-band for overlapped (bright) galaxies (～3700)Not necessarily true for all SDSS galaxies

Log (L100μm/Lr)

PDF of the ratioFit to log-

normal PDF

νrLr

IRAS/SDSS galaxiesν10

0μmL 1

00μ

m

IRAS/SDSS galaxies

Mock simulation result

the observed anomaly is reproduced qualitatively, but too strong if the mean and standard deviation of the ratio L100μm/Lr of IRAS/SDSS are assumed

black dashed: w/o ΔAblack solid: with ΔA

galaxy number per extinction bin after extinction correction

galaxy number before extinction correction

after extinction correction

before correction

Ar+ΔA [mag]S g

al[d

eg-2

]

Based on the parameters fitted to IRAS/SDSS

An analytic model for surface densityPDF of extra extinction ΔA due to FIR flux of one galaxy

PDF of N galaxies per pixel given its expectation value N

Conditional PDF that a pixel with N galaxies has the total

extra extinction ΔA

)(1 AP Δ

)|( NNP

)()()( 110xAPxPdxAP N

A

N −Δ=Δ −

Δ

∫

)|()()()()(0

0

NNPAPAAAdA N

A

N

ΔΔ−ΩΔ=Ω′ ∫∑∞

=

correctionwith10correctionwithout10

)|()()()()(

)(

)(

00

A

AA

Npixel

A

Ngal NNPAPAANAdAN

Δ

−Δ

∞

=

×

×

ΔΩ

Δ−ΩΔ=′ ∫∑

γ

γ

)()()( AANAS galgal Ω′′=′⇒

mock simulationsvs. analytic model

Area and galaxy number after contamination

Fit to the observed anomaly using an analytic approximation model

Basic trend is well reproduced with the model1/5 of the mean ratio L100μm/Lr of IRAS/SDSS overlapped sample, maybe reasonable

Ar,SFD [mag]S g

al[d

eg-2

]

before correction

after correction

DR7 obs. vs. best-fit model

IRAS/SDSS overlapped sample

Model fit to DR7 Sgal(ASFD)

Mean of L100μm/Lr

stan

dard

dev

iati

on o

f L 1

00μ

m/L

r in

log-

nor

mal

PD

F

ConclusionsAn anomaly in the SFD Galactic extinction map is confirmed with SDSS DR7 analysisThe anomaly would be due to FIR emission from galaxies

Implications on precision cosmology remain to be studied

More accurate Galactic extinction map with higher angular resolution AKARI IR map and/or Planck IR/temperature map